A major social media corporation recently allowed marketing campaigns promoting a controversial artificial intelligence tool directly to male users across its platforms.
The AI tool offered the ability to digitally strip high-profile American political figures, triggering intense backlash over online safety standards and automated moderation. As details emerge about how these promotions bypassed platform rules, questions are mounting over who approved the campaign and how far the technology was allowed to go.
Explicit AI Ads Appeared on Meta Despite Its Policies
Advertisements for an AI tool that nudged users into creating explicit clips of top American political figures recently slipped past Meta's safety filters, directly violating its rules against sexually explicit material. The slip-up highlights ongoing struggles within the tech giant to block campaigns promoting software that can generate non-consensual imagery across its networks.
Operating under the banner of an 'AI image styler', Kromix pitched itself to male audiences through video ads boasting real-world figures and zero creative limits. One clip reviewed by WIRED features a woman bearing a striking likeness to a well-known American leader before a presidential backdrop, captioned 'What if she moved?'
The video shows the figure winking before abruptly switching to explicit footage featuring the same face.
Before WIRED reached out for comment, it was reported that Apple's App Store hosted Kromix, which asked paying customers to upload photos to generate scenes involving 'bedroom rape' or 'Disney love'. The application's media library similarly features explicit generated clips depicting women dressed in Spider-Man costumes.
According to 9to5Mac, Apple pulled the software from its store shortly after being contacted by WIRED, explaining that Kromix seemingly introduced forbidden features only after clearing the initial approval process. Clarifying its stance, the company stated:
'We have always strictly prohibited apps designed to generate, distribute, or consume pornography. [...] "Nudification" apps are against our App Review Guidelines and we aggressively work to identify and remove apps—as well as the developer accounts behind them—that attempt to evade our processes and policies.'
AI Promotions Exclusively Targeted Male Users
According to Meta's ad library, the promotions were directed purely at male users under the slogan 'Unleash Next-gen AI Magic' and were pulled only after WIRED raised questions. The account behind the push launched 32 promotions in total, which remained active for anywhere between five and 46 hours, though most logged 10 views or fewer.
Apple continues to face difficulties keeping similar applications off its platform, despite rules forbidding non-consensual intimate imagery, nudification and explicit material. San Francisco's city attorney issued a cease-and-desist demand to the firm earlier this summer, insisting on the takedown of 13 specific applications and urging an end to 'aiding and abetting' the distribution of explicit deepfakes.
Digital Abuse Disproportionately Impacts Female Politicians
Katie Paul, director of the Tech Transparency Project, the watchdog group that spotted the promotions, pointed out that every detected nudification campaign specifically targets and sexualises women using synthetic imagery. She noted that this pattern raises serious doubts about Meta's moderation strategy, particularly when explicit promotions featuring public figures actively circulate on its networks.
Female politicians and elected representatives have been among those targeted by non-consensual deepfakes. The vulnerability was laid bare in January when the Grok tool was used to produce roughly 3 million sexualised images, with public figures among those whose likeness were used.
